On the ability of chemical transport models to simulate the vertical structure of the N2O, NO2 and HNO3 species in the mid-latitude stratosphere

(1)

HAL Id: hal-00327987

https://hal.archives-ouvertes.fr/hal-00327987

Submitted on 28 Nov 2005

HAL is a multi-disciplinary open access

archive for the deposit and dissemination of sci-entific research documents, whether they are pub-lished or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers.

L’archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d’enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

On the ability of chemical transport models to simulate

the vertical structure of the N2O, NO2 and HNO3

species in the mid-latitude stratosphere

Gwenaël Berthet, Nathalie Huret, Franck Lefèvre, G. Moreau, C Robert,

Michel Chartier, L. Pomathiod, Michel Pirre, Valéry Catoire

To cite this version:

Gwenaël Berthet, Nathalie Huret, Franck Lefèvre, G. Moreau, C Robert, et al.. On the ability of chemical transport models to simulate the vertical structure of the N2O, NO2 and HNO3 species in the mid-latitude stratosphere. Atmospheric Chemistry and Physics Discussions, European Geosciences Union, 2005, 5 (6), pp.12373-12401. �hal-00327987�

(2)

ACPD

5, 12373–12401, 2005 Simulation of the vertical structure of N₂O, NO₂ and HNO₃ G. Berthet et al. Title Page Abstract Introduction Conclusions References Tables Figures J I J I Back Close Full Screen / Esc

Print Version Interactive Discussion

EGU

Atmos. Chem. Phys. Discuss., 5, 12373–12401, 2005 www.atmos-chem-phys.org/acpd/5/12373/

SRef-ID: 1680-7375/acpd/2005-5-12373 European Geosciences Union

Atmospheric Chemistry and Physics Discussions

On the ability of chemical transport

models to simulate the vertical structure

of the N

₂

O, NO

₂

and HNO

₃

species in the

mid-latitude stratosphere

G. Berthet1,2, N. Huret2, F. Lef `evre1, G. Moreau2, C. Robert2, M. Chartier2, L. Pomathiod2, M. Pirre2, and V. Catoire2

1

Service d’A ´eronomie, Institut Pierre-Simon Laplace, F-75252 Paris, France

2

Laboratoire de Physique et Chimie de l’Environnement, F-45071 Orl ´eans, France Received: 19 April 2005 – Accepted: 28 June 2005 – Published: 28 November 2005 Correspondence to: G. Berthet (g.berthet@damtp.cam.ac.uk)

(3)

ACPD

EGU

Abstract

In this paper we study the impact of the modelling of N₂O on the simulation of NO₂and HNO₃by comparing in situ vertical profiles measured at mid-latitudes with the results of the Reprobus 3-D CTM (Three-dimensional Chemical Transport Model) computed with the kinetic parameters from the JPL recommendation in 2002. The analysis of the 5

measured in situ profile of N₂O shows particular features indicating different air mass origins. The measured N₂O, NO₂ and HNO₃ profiles are not satisfyingly reproduced by the CTM when computed using the current 6-hourly ECMWF operational analysis. Improving the simulation of N₂O transport allows us to calculate quantities of NO₂and HNO₃ in reasonable agreement with observations. This is achieved using 3-hourly 10

winds obtained from ECMWF forecasts. The best agreement is obtained by constrain-ing a one-dimensional version of the model with the observed N₂O. This study shows that modelling the NO_y partitioning with better accuracy relies at least on a correct simulation of N₂O and thus of total NO_y.

1. Introduction

15

The partitioning of individual nitrogen species within reactive nitrogen family (NO_y) is a very important factor in the stratospheric chemistry of ozone. The photolysis and the photo-oxidation of N₂O is the main source of NO_yspecies in the stratosphere:

N₂O+ hv → N₂+ O(1_D) ₍₁₎

N₂O+ O(1D) → N₂+ O2 (2a)

20

→ 2NO (2b)

Reaction (2b) is the dominant path for the production of NO_xspecies (e.g. McElroy and McConnell, 1971). NO_x (NO + NO₂), the most reactive form of the NO_y compounds,

(4)

ACPD

EGU

are involved in the main processes controlling the ozone budget. They both account for a major part of ozone reduction in the middle and upper stratosphere through rapid catalytic cycles (e.g. McElroy et al., 1992; Osterman et al., 1997; Br ¨uhl et al., 1998) and moderate the ozone-destroying catalytic cycles involving halogen and hydrogen radicals (e.g. Salawitch et al., 1994; Wennberg et al., 1994). NO_xare chemically linked 5

to nitric acid HNO₃ which is the main reservoir of the NO_y family below ∼30 km and are primarily produced by photolysis of HNO₃. The partitioning of NO_y species, and in particular NO₂and HNO₃, is also affected by heterogeneous reactions on sulphate aerosols that convert active nitrogen to HNO₃ (e.g. McElroy et al., 1992; Fahey et al., 1993).

10

Nevertheless the quantitative understanding of the partitioning of the NO_y species in the models remains limited. Several studies have shown significant discrepancies between computed and observed partitioning of NO_y species depending on the loca-tion (mid- or high latitudes) and time (day or night, period of the year) (e.g. Sen et al., 1998; Gao et al., 1999; Payan et al., 1999; Wetzel et al., 2002; Stowasser et al., 2002, 15

2003). Observations of the various NO_yspecies such as HNO₃ and NO₂are not sat-isfyingly reproduced by chemistry-transport models at mid-latitudes with in particular recurrent underestimations of NO₂mixing ratios (Wetzel et al., 2002; Stowasser et al., 2003). The different compounds among the NO_yfamily are also represented as ratios to separate dynamical from chemical effects on the NO_ypartitioning and to reduce the 20

influence of missing amounts of total NO_y in the models. Simulations of these ratios generally appear to be more satisfactory (e.g. Kondo et al., 2000; Wetzel et al., 2002) even though Stowasser et al. (2003) have presented NO₂/NO_y ratios strongly under-estimated by model results in the lower stratosphere. This seems to demonstrate that dynamical effects constitute the major part of the discrepancies with observations in 25

the middle-stratosphere.

The SPIRALE (French acronym for “Spectroscopie InFrarouge par Absorption de Lasers Embarqu ´es”) balloon-borne instrument is able to measure in situ profiles of NO₂and HNO₃and of their precursor N₂O with a high vertical resolution. Such in situ

(5)

ACPD

EGU

measurements appear to be more able to characterize local processes than remote sensing observations using lines of sight of a few hundred of kilometres. We report here measurements of these species at mid-latitudes on 2 October 2002 in the frame of the first validation campaign of the ENVISAT satellite. In this paper we try to balance the dynamical and chemical effects on the modelling of the NO_y partitioning. First 5

we assess the influence of the 3-D winds used in the Reprobus CTM on its ability to reproduce the NO_y partitioning. Then a one-dimensional (1-D) version of the model is used to test the impact of a correct initialisation of N₂O on the simulation of the chemistry controlling the NO_y partitioning. It is the first modelling study of the in situ simultaneous observations by SPIRALE of NO₂and HNO₃with a CTM.

10

2. The SPIRALE balloon-borne instrument

A detailed description of the instrumental characteristics of SPIRALE and of its operat-ing mode can be found in Moreau et al. (2005). To summarize, the SPIRALE balloon-borne instrument performs in situ simultaneous measurements of several long-lived and short-lived chemical species from the tropopause up to 40 km. It uses six tunable 15

salt laser diodes in the mid-infrared domain (3 µm to 10 µm). The six laser diodes are cooled and the beams are injected into a multipass Heriott cell located under the gondola. The cell (3.5 m) is deployed during the flight above the tropopause. Eighty six reflections occur, giving a 300 m optical path. Species concentrations are retrieved from direct absorption, by fitting experimental spectra with spectra calculated using HI-20

TRAN 2001 database (Rothman et al., 2003). The instrument provides measurements with a vertical resolution of a few meters.

SPIRALE is used routinely, in particular as part of European and satellite valida-tion campaigns. On 2 October 2002, during the ENVISAT validavalida-tion campaign, the SPIRALE flight was conducted during the morning at mid-latitude from the Aire sur 25

l’Adour launch base (France, 43.7◦N, 0.3◦W). In our study we analyse the measure-ments of N₂O, HNO₃and NO₂profiles obtained during the ascent of the balloon around

(6)

ACPD

EGU

08:00 UT. The data used in this study are averaged over a vertical range of 250 m. Rel-ative uncertainties on concentrations are then 3% for N₂O, and 7% for HNO₃and NO₂.

3. Model calculations

The Reprobus chemical-transport model (Lef `evre et al., 1994, 1998) has been widely used in previous studies of the stratospheric chemistry (e.g. Lef `evre et al., 1994; Deniel 5

et al., 1998; Hoppel et al., 2002). The model is designed to perform annual simulations as well as detailed process studies. It computes the evolution of 55 species by means of about 160 photolytic gas-phase and heterogeneous reactions, with a time step of 15 minutes in this study. A semi-lagrangian code transports 40 species or chemical fami-lies, typically long-lived tracers but also more unstable compounds which may have a 10

long lifetime in darkness. Kinetics parameters used in the present study are based on the most recent data in general taken from Sander et al. (2003). The new laboratory measurements of photodissociation cross-sections of HO₂NO₂both in the UV (Knight et al., 2002) and in the near IR (Roehl et al., 2002) have been included in the photodis-sociation calculations. The heterogeneous chemistry module includes reactions on 15

liquid aerosols. Their surface area densities are inferred from SAGE II measurements (Thomason et al., 1997).

Reprobus was integrated from 1 April 2002 to 15 October 2002. The model extends from the surface up to 0.1 hPa on 42 levels, resulting in a vertical resolution of about 1.3 km in the lower stratosphere. The horizontal resolution is 2◦latitude by 2◦longitude. 20

Temperature, winds and ground pressure are specified from the three-dimensional (3-D) European Centre for Medium-Range Weather Forecast (ECMWF) meteorological data. The ozone field was initialized on 1 April 2002 from the ECMWF ozone analysis. Other species were initialized from an April zonal mean computed from a long-term simulation of Reprobus.

(7)

ACPD

EGU

4. Comparisons between SPIRALE measurements and 3-D calculations

4.1. N₂O profile 4.1.1. Measurement

The in situ profile of N₂O measured by SPIRALE (Fig. 1) presents different features which suggest that the instrument observed air masses of different origins. In particular 5

a 5 km layer with a minimum value of 50 ppbv at 27 km is clearly present between 23 and 29 km.

A detailed study based on the SPIRALE observations of N₂O and CH₄ has been made to investigate the dynamical state of the stratosphere during this flight (Huret et al., 20051). Using a combination of [CH₄]:[N₂O] correlations and potential vorticity 10

maps calculated by the MIMOSA contour advection model (Hauchecorne et al., 2002) this study shows a perturbed situation probably due to the vertical wind shear occurring during the first stage of the polar vortex formation. From the [CH₄]:[N₂O] correlations, mid-latitude air was diagnosed between about 16 and 23 km whereas the air masses sampled above 29 km mainly originated from tropical latitudes. This is consistent with 15

the location of the instrument with respect to the dynamical barrier existing in this re-gion: the MIMOSA model calculates potential vorticity values in the 43–50 pvu range at 550 K (about 22.5 km) and close to 149 pvu at 810 K (about 30 km) at the location of the measurement at 06:00 UT (Fig. 2a and b). Between 23 and 29 km the N₂O profile is clearly non-monotonic (Fig. 1) and the vertical profiles of both tracers are associ-20

ated with a significantly different [CH₄]:[N₂O] correlation with respect to Michelsen et al. (1998) correlation curves (Huret et al., 20051). This gives an indication that mixing processes have probably occurred, similarly to the mixing event processes occurring

1

Huret, N., et al.: On the vertical structure of the stratosphere at mid-latitude during the first stage of the polar vortex formation and in polar region in presence of a large mesospheric descent, J. Geophys. Res., in preparation, 2005.

(8)

ACPD

EGU

inside the Arctic vortex and across the vortex edge (Plumb et al., 2003; Konopka et al., 2004) or to the mixing of polar vortex air into middle latitudes detected from tracer-tracer scatter plots (Waugh et al., 1997). The analysis of such a mixing event is out of the scope of our paper.

Similar conclusions can be derived from the comparisons performed by Urban et 5

al. (2004) between SPIRALE measurements and the observations of the SMR instru-ment onboard the Odin satellite (e.g. Murtagh et al., 2002). This study shows that averaging Odin/SMR measurements over 27.5–42.5◦N (mostly tropical air) and 42.5– 55◦N (mostly mid-latitude air) gives N₂O amounts similar to those observed by SPI-RALE above 29 km and below 23 km, respectively. Between 23 and 29 km the N₂O 10

profile measured by SPIRALE is located between the two averaged profiles observed by Odin/SMR, suggesting a mixing event.

4.1.2. Model calculations

CTM calculations were driven using the 3-D 6-hourly ECMWF operational analysis for winds, temperature and ground pressure recurrently employed in modelling studies 15

(e.g. Lef `evre et al., 1998; Wetzel et al., 2002; Stowasser et al., 2003; Ricaud et al., 2005). Figure 1 compares the observed profile of N₂O to the model result obtained at the grid point closest to the measurement position. The model slightly underes-timates the observations below 20 km but discrepancies are mainly observed above 25 km: Reprobus overestimates SPIRALE measurements of N₂O above 30 km, and 20

does not reproduce the structure at 27 km. Other measurements-model comparisons have revealed overestimations of the calculated amounts of N₂O. N₂O mixing ratios measured by the MIPAS-B balloon-borne instrument at mid-latitudes during summer time are overestimated by the KASIMA 3-D CTM below 30 km (Wetzel et al., 2002). Comparisons between Reprobus and the Odin satellite measurements have shown 25

that the model overestimated the N₂O amounts observed by Odin/SMR at mid-latitudes in mid-September 2002 (Ricaud et al., 2005). Similar disagreements are observed with preliminary measurements of the MIPAS instrument on board the ENVISAT satellite

(9)

ACPD

EGU

for the same period (R. Runhke, personal communication). This notorious problem appears to be related to a too strong vertical transport above the Equator when us-ing ECMWF operational analysis winds which provide too large quantities of N₂O in the tropical middle stratosphere (Ricaud et al., 2005). Then these amounts reach mid-latitudes by the Brewer-Dobson circulation and quasi-horizontal transport from the 5

tropical latitudes (Holton et al., 1995). 4.2. NO_ypartitioning

We do not consider the diurnal variation of NO₂since it appears to be weak between the half time of SPIRALE measurements during the ascent of the balloon (08:00 UT at about 24 km) and the last measurement (08:30 UT at float altitude). Figure 3 shows the 10

comparisons between the observed profiles of NO₂and HNO₃and the model calcula-tions for a solar zenith angle of 70◦ at 08:00 UT. Significant underestimation is visible on the simulated profile of NO₂above 22 km (33, 40 and 22% lower than the observa-tion at 23, 27 and 31 km, respectively) where N₂O is overestimated by the model. The model underestimates the measured mixing ratios of HNO₃ above 27 km (12, 19 and 15

26% lower than the observation at 27, 28 and 31 km, respectively). Simulated HNO₃is overestimated below 22 km (51 and 23% higher than the observation at 18 and 20 km, respectively) where N₂O is underestimated by the model. For other altitudes Reprobus results stay within the error bars of the SPIRALE measurements.

One must examine whether such underestimation of the mixing ratios of these two 20

major species among the NO_y family is related to the ability of the model to calculate the total amounts of NO_y. Indeed 3-D CTMs may underestimate the total amount of NO_y at mid-latitudes, as shown by the comparison between the KASIMA CTM and observations by the MIPAS-B instrument (Wetzel et al., 2002), which affects the sim-ulation of the various NO_y species. This result is supported by recent comparisons 25

between different European models with preliminary MIPAS observations (R. Ruhnke, personal communication). Since SPIRALE does not measure all of the NO_y species (N₂O₅, ClONO₂, NO₃ and HO₂NO₂ are missing), we have reconstructed the in situ

(10)

ACPD

EGU

profile of total NO_y (hereafter NO_y*) using the compact N₂O-NO_y relations measured in different latitude bands. Figure 4 presents the NO_y* profile inferred from the 28– 46◦N correlation given by Michelsen et al. (1998). The profile from the 3–10◦N latitude band correlation is also represented above 29 km where air masses appear to mainly originate from low latitudes (see Sect. 4.1.1). N₂O and NO_yare anticorrelated for N₂O 5

mixing ratios larger than about 50 ppbv. Subsequently, the comparison between the reconstructed and simulated profiles of NO_y(Fig. 4) reveals a significant underestima-tion by Reprobus above 25 km where N₂O is overestimated by the model. Above 30 km the model particularly underestimates the NO_y* profile inferred from the 3–10◦N N₂ O-NO_y correlation. A slight overestimation is observed at some altitudes below 20 km 10

where N₂O is underestimated by the model. These two behaviours are the result of the discrepancies observed in Fig. 1. Any problem in the modelling of N₂O has a direct impact on the calculated quantities of total NO_y. The transport of too large quantities of the N₂O precursor species at mid-latitudes stated above can result in a too weak production of NO_y after its transport from tropical latitudes (e.g. Ricaud et al., 2005) 15

which explains the low values for the upper part of the profile in Fig. 4.

To minimize the effect of the uncertainty in the total NO_y modelling we have com-pared the NO₂/HNO₃ratios rather than individual profiles (Fig. 5). In that case better results are obtained: the NO₂/HNO₃ratio is well reproduced by Reprobus above 27 km and between 17 and 22 km. However an underestimation can be seen between 22 20

and 27 km (34 and 31% lower than the observation at 23 and 27 km, respectively) as a result of NO₂ underestimation. From the NO₂/HNO₃ ratio we can see in this alti-tude range that the particular structure attributed to a mixing event affects our ability to reproduce the observed NO_ypartitioning. We conclude from Fig. 5 that the NO_y parti-tioning seems to be correctly reproduced by the CTM at all altitudes except at those of 25

the particular vertical structure. Note that the strong peaks of the NO₂/HNO₃ ratio at 15 and 16.5 km cannot be reproduced by the model which could be due to its too low vertical resolution.

(11)

ACPD

EGU

The modelling of N₂O relies mostly on transport. Thus, one can expect that a better simulation of the transport of N₂O can improve the agreement between measurements and calculations of the NO₂and HNO₃quantities.

5. Sensitivity to the wind fields

5.1. 6-hourly and 3-hourly ECMWF winds 5

In the work of Legras et al. (2005) in situ measurements of N₂O performed from the ER-2 aircraft are compared both to Reprobus 3-D outputs (computed at the time/positions of the flight track) and to reconstructed trajectories. Significant discrepancies are ob-served between the CTM and the reconstructed trajectories when computed with 6-hourly ECMWF operational analysis. These are characterized by strong variations and 10

unrealistic structures apparently as a result of high-frequency fluctuations in the winds. A dramatic improvement is obtained by Legras et al. (2005) when using 3-hourly analy-sis to compute air mass trajectories and the resulting N₂O distribution. We have tested this result by performing a second simulation using 3-hourly wind fields obtained by interleaving 3H and 9H forecasts between the ECMWF 6-hourly operational analysis. 15

Figure 6 compares N₂O zonal means calculated by Reprobus when driven by 6-hourly and 3-hourly winds on 2 October 2002, at the time of SPIRALE observations. It can be seen that using 3-hourly winds clearly leads to more realistic N₂O values in the upper stratosphere, as a result of a slower circulation and reduced vertical diffusion. This result is consistent with the analysis of Legras et al. (2005). It also supports the work 20

of Stohl et al. (2004) who show that forecast winds are less diffusive than operational analysis.

5.2. 3-D CTM calculations using 3-hourly winds

SPIRALE observations are now compared to two additional Reprobus 3-D CTM sim-ulations using different wind sources: 1) a simulation considering 3-hourly wind fields, 25

(12)

ACPD

EGU

interleaving operational analysis at 00:00, 06:00, 12:00, and 18:00 UT with forecasts valid at 03:00, 09:00, 15:00 and 21:00 UT, as in the work of Legras et al. (2005) (here-after named Reprobus-3h-ope-for). 2) a simulation only considering forecasts at 3-h interval, based on the analysis at 00:00 and 12:00 UT (hereafter Reprobus-3h-for). The “baseline” simulation using 6-hourly operational analysis is referred as Reprobus-ope 5

in the following. Initialising the transport every 3 h has clearly allowed us to model more accurately the vertical profile of N₂O tracer species at the time of the SPIRALE measurement (Fig. 7). Below 20 km the new N₂O profile is characterized by slightly higher values with a better agreement with Reprobus-3h-for. Above 20 km the N₂O quantities are decreased towards the profile obtained from the operational analysis. A 10

slight underestimation of the measurements is observed for the upper part of the profile but the decrease of N₂O amounts as altitude increases appears to be more consistent than the Reprobus-ope simulation results. However, the vertical structure observed by SPIRALE around 27 km cannot be reproduced with the new simulations.

The Reprobus-3h-ope-for and Reprobus-3h-for simulations result in enhanced total 15

mixing ratios of NO_yabove 20 km where N₂O calculated values are lower in comparison with the typical Reprobus-ope simulation (Fig. 8). Below these altitudes, the increased simulated amounts of N₂O give lower NO_y mixing ratios in better agreement with the NO_y* in situ profile deduced from Michelsen et al. (1998)’s correlation. Above 30 km however, the modelled total amounts of NO_y from the two additional simulations do 20

not appear to be sufficient when compared to the two NO_y* profiles (cf. Sect. 4.2). This could be considered as a limitation of the use of the Michelsen et al. (1998)’s correlation curve between N₂O and NO_y.

NO₂ and HNO₃ profiles calculated by the 3-hourly wind simulations have increased mixing ratios which are in better quantitative agreement with SPIRALE observations 25

(Fig. 9). For example, NO₂ quantities calculated with the Reprobus-3h-for simulation above 22 km are increased by 31, 22 and 12% at 23, 27 and 31 km, respectively, in comparison with the Reprobus-ope simulation (Fig. 9a). Nevertheless, underestima-tion is still visible around the 27-km structure and also above 30 km as a result of the

(13)

ACPD

EGU

apparent underestimation of total NO_y* quantities (27 and 12% lower than the obser-vation at 27 and 31 km, respectively). Below 22 km the disagreement remain roughly similar between the three model experiments. For HNO₃, most of the Reprobus-3h-for simulation values are located within the error bars of the SPIRALE profile (Fig. 9b), even though some fine specific structures cannot be reproduced, such as at 21 km 5

(36% higher than the observation). Slight improvement is obtained in comparison with the Reprobus-3h-ope-for experiment below 22 km where Reprobus-3h-for simulation gives the best agreement with the NO_y* in situ profile. The results of the two 3-hourly simulations of the NO₂/HNO₃ratio fall now within the error bars of the in situ observa-tion (Fig. 9c) which shows that the mid-latitude partiobserva-tioning between these two species 10

can be modelled correctly provided that the simulated NO_ytotal amounts are well taken into account.

Ignoring the specific problem of mixing, the Reprobus 3-D CTM underestimates the NO₂and HNO₃amounts at mid-latitudes due to the underestimation of total NO_ywhich is in turn the result of transport uncertainties in the model that affect the simulation of 15

N₂O. ECMWF 6-hourly operational analysis seem to be characterized by some noisy and unrealistic features which are not present in the forecasts (Legras et al., 2005) and apparently affecting the transport calculations in Reprobus and in other CTMs. The strict reasons for the significant improvements of the NO_y comparisons using the 3-hourly winds are still unclear. Nevertheless it is thought that using 3-3-hourly wind data 20

gives a better sampling for the interpolation of the positions between two consecutive wind records in Reprobus.

However discrepancies between the SPIRALE measurements and the model results in Fig. 9 still remain especially for NO₂above 30 km as a result of the underestimation of the total amounts of NO_y* at these altitudes. Recent studies have shown that satisfy-25

ing comparisons can be obtained between observations and model calculations when those are constrained by observations of total NO_y (e.g. K ¨ull et al., 2002). This points out the necessity of an adequate simulation of total amounts of NO_y to reproduce the partitioning of the individual species.

(14)

ACPD

EGU

6. 1-D calculations

To get rid of the influence of the problem of transport and hence of the missing quan-tities of total NO_y, we have used the Reprobus model in its one-dimensional form. Constraining 1-D-Reprobus calculations by observed total NO_y has significantly im-proved the modelled partitioning of NO_y species compared to balloon measurements 5

in the summer polar stratosphere (Dufour et al., 2005). Similar conclusions have been obtained from comparisons between balloon observations and lagrangian simulations in January 1999 inside a warm Arctic vortex (Stefan Bausch, private communication). In the present study the 1-D-Reprobus initialisation is constrained by the N₂O profile measured by SPIRALE and by the corresponding NO_y* profiles. However the N₂O-NO_y 10

correlation curves from Michelsen et al. (1998) have been determined on given latitude bands and it is not straightforward to choose the adequate correlation values that have to be considered above 29 km where air seems to originate from low latitudes. Two simulations are performed in our study. The first one (referred as Reprobus-1D-midlat) is constrained with the NO_y* profile inferred from the N₂O-NO_ycorrelations in the 28– 15

46◦N latitude band. In the second simulation (hereafter Reprobus-1D-midlat-trop) the computed NO_y* profile is defined by the N₂O-NO_y correlation values in the 28–46◦N and 3–10◦N latitude ranges below and above 29 km, respectively (cf. Sect. 4.2). These two model experiments allow us to test the sensitivity of the choice of the N₂O-NO_y correlation law on the modelled partitioning of NO₂and HNO₃. The other species are 20

initialised from the 3-D-Reprobus simulation as in the work of Dufour et al. (2005). For both experiments the calculated mixing ratios of NO₂and HNO₃are close to the measurements (Fig. 10). For NO₂ significant improvement is obtained above 22 km: the Reprobus-1D-midlat experiment provides values that are only 17% and 3% lower than the SPIRALE profile at 23 and 27 km, respectively. Above 29 km the measured 25

quantities lie between the results of the two simulations. An average of the results of both simulations is in perfect agreement with the observation of NO₂ at these alti-tudes. Similar conclusions can be derived from the HNO₃profile which is satisfyingly

(15)

ACPD

EGU

simulated for the whole altitude range (24% and 10% lower than the observation at 18 and 20 km, respectively, 6% and 24% higher than the observation at 23 km and 28 km, respectively) except in the layer around 27 km (41% higher than the observation at 27 km). The calculation of the NO₂/HNO₃ ratio is not significantly improved between 22 and 27 km (22% and 31% lower than the observation at 23 km and 27 km, respec-5

tively), the computed ratio being underestimated at 27 and 28 km as a result of HNO₃ overestimation. Taking into account the N₂O-NO_y correlation uncertainty of Michelsen et al. (1998) the 1-D-simulation cannot reproduce simultaneously NO₂ and HNO₃ in the particular layer of a possible previous mixing event. Actually the correlation curves of Michelsen et al. (1998) determined on given latitude bands cannot account for a full 10

description a N₂O-NO_ycorrelation within a layer containing mixed air from tropical and mid- latitudes.

7. Conclusions

We have used in situ observations of N₂O, NO₂ and HNO₃ by the SPIRALE balloon-borne instrument to test our ability to reproduce the observed NO_y partitioning at mid-15

latitudes. In a first part we have used the Reprobus 3-D CTM computed with the most recent kinetics data from the JPL recommendation and driven by the meteorological parameters from ECMWF. Discrepancies appear between the simulation and the mea-surement of N₂O when transport calculations are conducted with the ECMWF 6-hourly operational analysis widely used in the scientific community. This affects the mod-20

elling of NO₂and HNO₃ with in particular an underestimation of the NO₂/HNO₃ ratio. The comparisons for N₂O are improved at all altitudes when transport calculations are driven by 3-hourly wind data. This result is consistent with the analysis of Legras et al. (2005) who shows much better agreement between ER-2 in situ observations and distributions of N₂O calculated using such 3-hourly wind fields. From this improved 25

simulation, more realistic N₂O, NO₂, and HNO₃ profiles are calculated by Reprobus. However, the vertical feature observed by SPIRALE around 27 km in the N₂O profile

(16)

ACPD

EGU

cannot be reproduced by the 3-D CTM. This feature is attributed to a previous mixing event which is beyond the capabilities of the model at its current horizontal and vertical resolution.

The best results regarding the partitioning of NO_y are obtained by constraining the 1-D version of model with the observed profile of N₂O and the corresponding total 5

NO_y deduced from established N₂O-NO_y correlations. In these conditions, our model simulations show that the observed mid-latitude NO₂/HNO₃ ratio can be reproduced with excellent agreement at all stratospheric altitudes. Modelling the NO_y partitioning with the best accuracy relies at least on a satisfying simulation of N₂O and thus of total NO_y. This result is consistent with the conclusions of Dufour et al. (2005) who 10

constrained their model for the Arctic summer period with the sum of the measured NO_y species. It also suggests that discrepancies found between observations and modelling of the NO_ypartitioning at mid-latitudes mainly originate from transport effects rather than general knowledge of the stratospheric chemistry.

The problem of transport calculation may also affect the modelling of diabatic descent 15

inside the polar vortex where disagreements appear when comparisons are performed between models and satellite observations (e.g. Ricaud et al., 2005). A next step of our work is to make comparisons at a global scale and for long time series using satellite data such as those of ODIN and ENVISAT.

Acknowledgements. We are grateful to all the members of the CNES launching team at Aire

20

sur l’Adour. This work was funded in part by the European Commission under the TOPOZ-III project (contract EVK2-CT-2001-00102).

References

Br ¨uhl, C., Crutzen, P. J., and Grooß, J.-U.: High-latitude, summertime NO_x activation and seasonal ozone decline in the lower stratosphere: Model calculations based on observations

25

(17)

ACPD

EGU

Deniel, C., Pommereau, J. P., Bevilacqua, R. M., and Lef `evre, F.: Arctic chemical depletion during the 1994–95 winter deduced from POAM II satellite observations and the Reprobus-3D model, J. Geophys. Res, 103, 19 231–19 244, 1998.

Dufour, G., Payan, S., Lef `evre, F., Eremenko, M., Butz, A., Jeseck, P., T ´e, Y., Pfeilsticker, K., and Camy-Perret, C.: 4-D comparison method to study the NO_ypartitioning in summer polar

5

stratosphere – Influence of aerosol burden, Atmos. Chem. Phys., 5, 919–926, 2005,

SRef-ID: 1680-7324/acp/2005-5-919.

Fahey, D. W., Kawa, S. R., Woodbridge, E. L., et al.: In situ measurements constraining the role of sulphate aerosols in mid-latitude ozone depletion, Nature, 363, 509–514, 1993.

Gao, R. S., Fahey, D. W., Del Negro, L. A., et al.: A comparison of observations and model

sim-10

ulations of NO_x/NO_yin the lower stratosphere, Geophys. Res. Lett, 26, 1153–1156, 1999. Hauchecorne, A., Godin, S., Marchand, M., Hesse, B., and Souprayen, C.: Quantification

of the transport of chemical constituents from the polar vortex to midlatitudes in the lower stratosphere using the high-resolution advection model MIMOSA and effective diffusivity, J. Geophys. Res., 107(D20), 8289, doi:10.1029/2001JD000491, 2002.

15

Holton, J. R., Haynes, P. H., McIntyre, M. E., Douglass, A., Rood, R., and Pfister, L.: Stratosphere-troposphere exchange, Rev. Geophys., 33(4), doi:10.1029/95RG02097, 1995. Hoppel, K., Bevilacqua, R., Nedohula, G., Deniel, C., Lef `evre, F., Lumpe, J., Fromm, M.,

Ran-dall, C., Rosenfield, J., and Rex, M.: POAM III observations of arctic ozoneloss for the 1999/2000 winter, J. Geophys. Res., 107, doi:10.1029/2001JD000476, 2002.

20

Knight, G., Ravishankara, A. R., and Burkholder, J. B.: UV absorption cross sections of HO₂NO₂between 343 and 273 K, Phys. Chem. Phys., 4, 1732–1437, 2002.

Kondo, Y., Sugita, T., Koike, M., Kawa, S. R., Danilin, M. Y., Rodriguez, J. M., Spreng, S., Golinger, K., and Arnold, F.: Partitioning of reactive nitrogen in the mid-latitude lower strato-sphere, J. Geophys. Res., 105, 1417–1424, 2000.

25

Konopka, P., Steinhorst, H.-M., Grooß, J.-U., et al.: Mixing and ozone loss in the 1999–2000 Arctic vortex: Simulations with the three-dimensional Chemical Lagrangian Model of the Stratosphere (ClaMS), J. Geophys. Res., 109, D02315, doi:10.1029/2003JD003792, 2004. K ¨ull, V., Riese, M., Tie, X., Wiemert, T., Eidmann, G., Offermann, D., and Brasseur, G.:

NO_y partitioning and aerosol influences in the stratosphere, J. Geophys. Res., 107, 8183,

30

doi:10.1029/2001JD001246, 2002.

Lef `evre, F., Brasseur, G. P., Folkins, I., Smith, A. K., and Simon, P.: Chemistry of the 1991– 1992 stratospheric winter : Three-dimensional model simulations, J. Geophys. Res., 99,

(18)

ACPD

EGU

9183–8195, 1994.

Lef `evre, F., Figarol, F., Carslaw, K., and Peter, T.: The 1997 Arctic ozone depletion quantified from three-dimensional model simulations, Geophys. Res. Lett., 25, 2425–2428, 1998. Legras, B., Pisso, I., Berthet, G., and Lef `evre, F.: Variability of the Lagrangian turbulent

diffu-sivity in the lower stratosphere, Atmos. Chem. Phys., 5, 1605–1622, 2005,

5

SRef-ID: 1680-7324/acp/2005-5-1605.

McElroy, M. B. and McConnell, J. C.: Nitrous oxide: A natural source of stratospheric NO, J. Atmos. Sci., 28, 1095–1098, 1971.

McElroy, M. B., Salawitch, R. J., and Minschwaner, K.: The changing stratosphere, Planet. Space Sci., 40, 373–401, 1992.

10

Michelsen, H. A., Manney, G. L., Gunson, M. R., and Zander, R.: Correlations of stratospheric abundances of NO_y, O₃, N₂O, and CH₄ derived from ATMOS measurements, J. Geophys. Res., 103, 28 347–28 359, 1998.

Moreau, G., Robert, C., Catoire, V., Chartier, M., Camy-Peyret, C., Huret, N., Pirre, M., and Pomathiod, L.: A multi-species in situ balloon-borne instrument with six diode laser

spec-15

trometers, Appl. Opt., 44(28), 1–18, 2005.

Murtagh, D., Frisk, U., Merino, F., et al.: An overview of the Odin atmospheric mission, Can. Journ. of Phys., 80, 309–319, 2002.

Osterman, G. B., Salawitch, R. J., Sen, B., et al.: Balloon-borne measurements of stratospheric radicals and their precursors: Implications for production and loss of ozone, Geophys. Res.

20

Lett, 24, 1107–1110, 1997.

Payan, S., Camy-Peyret, C., Jeseck, P., Hawat, T., Pirre, M., Renard, J.-B., Robert, C., Lef `evre, F., Kanzawa, H., and Sasano, Y.: Diurnal and nocturnal distribution of stratospheric NO₂from solar and stellar occultation measurements in the Arctic vortex: comparison with models and ILAS satellite measurements, J. Geophys. Res., 104, 21 585–21 593, 1999.

25

Plumb, R. A., Heres, W., Neu, J. L., Mahowald, N. M., del Corral, J., Toon, G. C., Ray, E., Moore, F., and Andrews, A. E.: Global tracer modelling during SOLVE: High latitude descent and mixing, J. Geophys. Res., 108(D5), 8309, doi:10.1029/2001JD001023, 2003.

Ricaud, P., Lef `evre, F., Berthet, G., et al.: Polar vortex evolution during the 2002 Antarc-tic major warming as observed by the Odin satellite, J. Geophys. Res., 110, D05302,

30

doi:10.1029/2002JD005018, 2005.

Roehl, C. M., Nizkorodov, S. A., Zhang, S. A., et al.: Photodissociation of Peroxynitric Acid in the Near-IR, J. Phys. Chem., 106, 3766-3772, 2002.

(19)

ACPD

EGU

Rothman, L. S., Barbe, A., Benner, D. C., et al.: The HITRAN molecular spectroscopic database: edition of 2000 including updates through 2001, J. Quantitative Spect. Radiative Transfer, 82(1-4), 5–44, 2003.

Salawitch, R. J., Wofsy, S. C., Wennberg, P. O., et al.: The distribution of hydrogen, nitro-gen, and chlorine radicals in the lower stratosphere: Implications for changes in O₃ due to

5

emission of NO_yfrom supersonic aircraft, Geophys. Res. Lett., 21, 2547–2550, 1994. Sander, S. P., Friedl, R. R., Ravishankara, A. R., et al.: Chemical Kinetics and Photochemical

Data for use in atmospheric studies, Evaluation number 14, JPL Publication 02-25, 2003. Sen, B., Toon, G. C., Osterman, G. B., Blavier, J.-F., Margitan, J. J., Salawitch, R. J., and

Yue, G. K.: Measurements of reactive nitrogen in the stratosphere, J. Geophys. Res., 103,

10

3571–3585, 1998.

Stohl, A., Cooper, O., and James, P.: A cautionary note on the use of meteorological analysis fields for quantifying atmospheric mixing, J. Atmos. Sci., 61, 1446–1453, 2004.

Stowasser, M., Oelhaf, H., Ruhnke, R., Wetzel, G., Friedl-Vallon, F., Kleinert, A., Kouker, W., Lengel, A., Maucher, G., Nordmeyer, H., Reddmann, Th., Trieschmann, O., v.

Clar-15

mann, T., and Fischer, H.: A characterization of the warm 1999 Arctic winter by observa-tions and modeling: NO_y partitioning and dynamics, J. Geophys. Res., 107(D19), 4376, doi:10.1029/2001JD001217, 2002.

Stowasser, M., Oelhaf, H., Ruhnke, R., Kleinert, A., Wetzel, G., Friedl-Vallon, F., Kouker, W., Lengel, A., Maucher, G., Nordmeyer, H., Reddmann, Th., and Fischer, H.: The variation

20

of short-lived NOy species around sunrise at mid-latitudes as measured by MIPAS-B and calculated by KASIMA, Geophys. Res. Lett, 30, 1432, doi:10.1029/2002GL016727, 2003. Thomason, L. W., Poole, L. R., and Deshler, T. R.: A global climatology of stratospheric aerosol

surface area density as deduced from SAGE II: 1984–1994, J. Geophys. Res., 102, 8967– 8976, 1997.

25

Urban, J., Lauti ´e, N., Le Flochmo ¨en, E., et al.: Odin/SMR Limb Observations of Stratospheric Trace gases: Validation of N₂O, J. Geophys. Res., 110, D09301, doi:10.1029/2004JD005394, 2004.

Waugh, D. W., Plumb, R. A., Elkins, J. W., et al.: Mixing of polar vortex air into middle latitudes as revealed by tracer-tracer scatterplots, J. Geophys. Res., 102, 13 119–13 134, 1997.

30

Wennberg, P. O., Cohen, R. C., Stimpfle, R. M., et al.: Removal of stratospheric O₃by radicals: In situ measurements of OH, HO2, NO2, ClO and BrO, Science, 266, 398–404, 1994.

Wetzel, G., Oelhaf, H., Ruhnke, R., Friedl-Vallon, F., Kleinert, A., Kouker, W., Maucher, G.,

(20)

ACPD

EGU

Reddmann, T., Seefeldner, M., Stowasser, M., Trieschmann, O., Von Clarmann, T., and Fischer, H.: NO_ypartitioning and budget and its correlation with N₂O in the Arctic vortex and in summer midlatitudes in 1997, J. Geophys. Res., 107(D16), doi:10.1029/2001JD000916, 2002.

(21)

ACPD

EGU

Fig. 1. In situ volume mixing ratios (VMRs) of N₂O measured by the SPIRALE instrument on 2 October 2002 at mid-latitudes (black line) and compared to the VMRs simulated by the Reprobus 3-D CTM at the time of SPIRALE observations (08:00 UT) and obtained at the grid point closest to the measurement position (red line). The model is computed using ECMWF operational analysis data. For more clarity the measurement error bars are represented every 100 m.

(22)

ACPD

EGU

Fig. 2. Potential Vorticity (PV) isocontours calculated by the MIMOSA high-resolution advection

model at 06:00 UT on 2 October 2002 at: (a) 550 K (about 22.5 km); and (b) 810 K (about

30 km). The red regions denote air masses from tropical latitudes. The black cross denotes the SPIRALE instrument location.

(23)

ACPD

EGU

Fig. 3. Comparison between measured vertical profiles of NO₂ (a) and HNO₃ (b) on 2

Oc-tober 2002 at mid-latitudes (black lines) and profiles simulated by the Reprobus 3-D CTM at 08:00 UT and a solar zenith angle of about 70◦ (red lines). The model is computed using ECMWF operational analysis data.

(24)

ACPD

Print Version Interactive Discussion EGU

Fig. 4. Vertical profiles of reconstructed NO_y* compared to the Reprobus model output on 2 October 2002 (red line). Black curve: NO_y* profile inferred from the N₂O-NO_y correlation determined by Michelsen et al. (1998) in the 28–46◦N latitude band. Orange curve: NO_y* profile inferred from the correlation valid in the 3–10◦N latitude band (only represented above 29 km where the air seems to mostly originate from low latitudes).

(25)

ACPD

EGU

Fig. 5. NO₂/HNO₃ratio measured by SPIRALE (black line) and calculated by Reprobus 3-D CTM at 08:00 UT (red line). The model is driven by ECMWF 6-hourly operational analysis.

(26)

ACPD

Print Version Interactive Discussion EGU      

Fig. 6. N₂O zonal means simulated by Reprobus on 2 October 2002 at 08:00 UT. (a) Results

obtained using 6-hourly ECMWF operational analysis data;(b) Results obtained using 3-hourly

(27)

ACPD

EGU

Fig. 7. Same as Fig. 1 but the comparison is made with the results of three model simulations.

Red line: 6-hourly operational analysis (Reprobus-ope); green line: 3-hourly data obtained by interleaving operational analysis with forecasts (Reprobus-3h-ope-for, see text); blue line: 3-hourly forecasts available since 2002 (Reprobus-3h-for).

(28)

ACPD

Fig. 8. Same as Fig. 4 but the comparison is made with the results of the three model

simula-tions. Red line: 6-hourly operational analysis (Reprobus-ope); green line: 3-hourly data from operational analysis and forecasts (Reprobus-3h-ope-for, see text); blue line: 3-hourly forecasts (Reprobus-3h-for).

(29)

ACPD

Fig. 9. NO₂ (a) and HNO₃ (b): same as Fig. 3 but the comparison is made with the

re-sults of the three model simulations (Reprobus-ope, ope-for and Reprobus-3h-for). NO₂/HNO₃ ratio(c): same as Fig. 5 but the comparison is made with the results of the

three model simulations.

(30)

ACPD

Fig. 10. NO₂(a), HNO₃(b) and NO₂/HNO₃ratio(c) measured by SPIRALE (black curve) and

simulated by Reprobus. The model is here constrained by the measured profile of N₂O and the NO_y* deduced from the N₂O-NO_ycorrelation. Blue curve: Reprobus-1D-midlat experiment, NO_y* profile inferred from the correlation valid in the 28–46◦N latitude band. Orange curve: Reprobus-1D-midlat-trop experiment, NO_y* profile inferred from the correlation valid in the 3– 10◦N latitude band, only shown above 29 km where the air seems to mostly originate from low latitudes.

On the ability of chemical transport models to simulate the vertical structure of the N<sub>2</sub>O, NO<sub>2</sub> and HNO<sub>3</sub> species in the mid-latitude stratosphere

HAL Id: hal-00327987

https://hal.archives-ouvertes.fr/hal-00327987

On the ability of chemical transport models to simulate

the vertical structure of the N2O, NO2 and HNO3

species in the mid-latitude stratosphere

Gwenaël Berthet, Nathalie Huret, Franck Lefèvre, G. Moreau, C Robert,

Michel Chartier, L. Pomathiod, Michel Pirre, Valéry Catoire

To cite this version:

ACPD

On the ability of chemical transport

models to simulate the vertical structure

of the N

2

O, NO

2

and HNO

3

species in the

mid-latitude stratosphere

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

ACPD

₂

₂

₃